|
Sunday, November 25, 2007
One puzzle facing many novice AdWords advertisers is - what is this {Keywords:} notation we often see colleagues using in ads? And, should I be using it?
Well, first up, I'm not about to get into a whole lot of exotic tricks that might help an affiliate bidding 5 cents on 50,000 different keywords to eke out a living based on slightly higher CTR's. Those individuals will be far more motivated than I am to run tests on exotic uses.
But what about the basics?
How it works: As many of you know, using the notation {KeyWord:Alternate Text Here} in, say, the title of the ad, will put whatever the user typed in as the title of the ad. So if they typed in "green snowboards" the title of your ad becomes Green Snowboards. If they type in something that is too long, contains awkward punctuation, or is otherwise unsuitable for an ad title, your alternate text (in this case it might have been Snowboards for Sale) would show up.
Why would you do this? I've been stunned on a least two occasions to hear experts telling large crowds of people that this method "always" "works." Note I put the scare quotes around "always" and works" because of course it doesn't always work, and you need to define what "works" means. (Or what "performs better" means.)
This is often most effective for lowball bidders who need high CTR above all other considerations. CTR's often do rise when ad titles and text seem customized to the user's exact need, but this situation is fluid.
Put simply, it's not a "yes or no" proposition, but just another testing element. An A/B test might test one fixed headline vs. one with dynamic keyword insertion. A multivariate test might include this test in the batch of other variables being tested.
And what are we testing? Usually it's an ROI-related measure such as cost per acquisition. Quite often, at least half the time in my experience, dynamic keyword insertion hurts ROI, because Pavlovian clicks are actually bad - they cost you money by being too attractive to impulsive surfers, while no more attractive or appropriate to actual buyers. Try to look past CTR's alone.
Exactly matching your copy with users' queries also leans heavily in the direction of treating the web merely like a commodity catalog ("I want two-by-fours! I want model #134454AA512!"), with no brand differentiation. While I do firmly believe that at this stage of evolution, our task if often 60-70% navigational and only 30-40% persuasive, that balance will tip over time, to 40% navigation and 60% brand identity and persuasion. Advertising is a persuasive medium that persuades integratively over time across multiple channels. Even AdWords isn't pure direct response.
And hey, if you've got a lot of tightly built ad groups, shouldn't the title match or closely match your keywords anyway? Yes it should!
So this method is typically a "cleanup on Aisle 7" type of fix for certain campaigns, or parts of campaigns, that you've built lazily or in the sense of "here's a bunch of miscellaneous extra stuff in the industry I'm working on."
The one-word advice that "always" "works": TEST.
Posted by
Andrew Goodman
View Posts by Category |
|
|
